A worrying statistic from World Health Organisation (WHO) indicated that approximately 1.3 million fatalities happen due to road accidents, worldwide. The development of driverless car system which could help the driver while maneuvering is a potential improvement in the future. In autonomous car features, traffic light is part of the object being recognized for its function on the road. Nevertheless, the process of detection may involve the problem to segregate it with the environment due to the colour related. The video stream from the camera is used as the input and the image of traffic light which are extracted use for the training processed. The ‘You Only Look Once’ (YOLO) method was used as object recognition which to spatially separated bounding boxes and associated class probabilities as a regression problem. Hue – Saturation – Value (HSV) colour space is the used to separates a colour-based image to a binary image (white responding to the colour filtered), which later being smoothed using a median filter to remove noise. Circle Hough transform technique is applied to detect the circle that found on images. The status of the traffic light can be interpreted by using the colour detected in the circle shape detected. The traffic light recognition accuracy is better at daytime with 99.68% vs nighttime at only 70.24%. The three colour detection for daytime also showing better accuracy at average of 95.96% vs 91.48% at nighttime. The detection rate for the circle shape also is much better than nighttime which at 94.5% vs 79.56%.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Traffic Light (Circle) Detection and Recognition Using YOLO and Image Processing Technique


    Weitere Titelangaben:

    Lect. Notes Electrical Eng.


    Beteiligte:
    Md. Zain, Zainah (Herausgeber:in) / Sulaiman, Mohd. Herwan (Herausgeber:in) / Mohamed, Amir Izzani (Herausgeber:in) / Bakar, Mohd. Shafie (Herausgeber:in) / Ramli, Mohd. Syakirin (Herausgeber:in) / Sani, Zamani Md. (Autor:in) / Saari, Mohd Iqbal Farez bin (Autor:in) / Izzudin, Tarmizi Ahmad (Autor:in)


    Erscheinungsdatum :

    09.03.2022


    Format / Umfang :

    11 pages





    Medientyp :

    Aufsatz/Kapitel (Buch)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    Deep Learning based smart traffic light system using Image Processing with YOLO v7

    Rangari, Avinash Padmakar / Chouthmol, Ashwini Ravindra / Kadadas, Chaitanya et al. | IEEE | 2022


    Intelligent Traffic Light System Using YOLO

    Sai Venu Prathap, K. / Srinivasulu Reddy, D. / Madhusudhan, S. et al. | Springer Verlag | 2022


    Yolo-Based Traffic Sign Recognition Algorithm

    Ming Li / Li Zhang / Linlin Li et al. | DOAJ | 2022

    Freier Zugriff

    A YOLO Based Approach for Traffic Light Recognition for ADAS Systems

    Mostafa, Miand / Ghantous, Milad | IEEE | 2022


    Traffic Signs Detection and Recognition Using A combination of YOLO and CNN

    Al Khafaji, Yasir Ali / El Abbadi, Nidhal K. | IEEE | 2022